Quote from: capn on August 13, 2020, 03:26:05 PM

Is this method so safe? For some reason, I'm not sure.

--------------------------
In a keyless system that does not have a key, but has a continuously changing set of encryption schemes, it is necessary to perform exactly the same task for absolutely every data packet with a volume of 256 bits.
Why?
Because for any and every 256-bit message, one unique encryption scheme is used (in fact, this is a unique set of encryption schemes and rules).

Consider attack resistance.
First. If the message contains only 10 data packets of 256 bits each, this is 10 times 2256 bits of information, then a brute force attack will have to be carried out absolutely on each data packet.
Mathematically, this means that with respect to the key encryption model, the task becomes more complicated as many times as there are data packets (256 bits each) a message contains.
Second. In contrast to the key encryption model, in a keyless system, the hypothetical positive result of a successful brute-force attack of any number of data packets (256 bits each) does not help to solve the problem of decrypting other data packets that make up this message.
Third. Thus, a rough search will have to be done for each data packet from the available set. If G is the minimum number of data packets, adding up which it is possible to unambiguously understand the open message, then the exhaustive search problem will look like this: it will be necessary to check 2 to the power (G * 256) options. The possibility of attacking such numbers needs no comment, it is utopia by definition for any high technological level of attackers.
Fourth. Any model of keyless encryption, technologically, must have the function of "encryption of silence", which simulates the exchange of cipher codes of open messages in this closed communication channel. If this function is there, therefore, you can use it as many times as necessary. This means that the number of packets that must be simultaneously decoded to understand an open message can be any large, regardless of the minimum size of the open message itself. How to solve the problem of breaking a cipher with such an additional condition? I can not imagine.